Open anfarr opened 2 years ago
Hi,
Thank you for pointing out the code mistake. Concerning the nextflow issue, I never tried to run APERO with nextflow and I don't have any idea about what is happening. I suppose that you used the same APERO command with singularity and nextflow ?
Simon
Hey, i wrote a small parser for the two functions, that is being run in the singularity container. But in essence the parameters being passed to APERO were the same. I will double check. For now i wrote a bash script that schedules the slurm jobs (running APERO in the singularity container on my files).
Edit: closed on accident
Anton
Hello Simon, first thank you for providing your program on github. I am currently trying to use your tool in a nextflow pipeline. So far the first function works as intended. However i had some problems with your second function "APERO_end_detection".
When using it with a ptt file, i realized that in line 509 the variable ptt had to be refactored to ptt_file
My first tests were successful in creating an output of the second function by simply running it in a singularity container on our cluster. Calls for APERO_end_detection were distributed via slurm to one node each with four threads respectively. However, when i do the same with nextflow and let it handle the slurm scheduling the majority of return values being generated by the APERO_end_detection calls are "NA".
The corresponding errors are:
X nodes produced errors; first error: missing value where TRUE/FALSE needed
where X is 1 < X <= "chosen number of threads".Maybe it is related to this problem https://stackoverflow.com/questions/16895848/results-of-workers-not-returned-properly-snow-debug. Although i am not sure why this is only the case when nextflow handles the scheduling.
Maybe you have an idea.